Professional Documents
Culture Documents
Principles of Hydrodynamic Lubrication: Fig. 1-2 Fig. 1-3
Principles of Hydrodynamic Lubrication: Fig. 1-2 Fig. 1-3
Principles of Hydrodynamic
Lubrication
4.1 INTRODUCTION
A hydrodynamic plane-slider is shown in Fig. 1-2 and the widely used hydro-
dynamic journal bearing is shown in Fig. 1-3. Hydrodynamic lubrication is the
fluid dynamic effect that generates a lubrication fluid film that completely
separates the sliding surfaces. The fluid film is in a thin clearance between two
surfaces in relative motion. The hydrodynamic effect generates a hydrodynamic
pressure wave in the fluid film that results in load-carrying capacity, in the sense
that the fluid film has sufficient pressure to carry the external load on the bearing.
The pressure wave is generated by a wedge of viscous lubricant drawn into the
clearance between the two converging surfaces or by a squeeze-film action.
The thin clearance of a plane-slider and a journal bearing has the shape of a
thin converging wedge. The fluid adheres to the solid surfaces and is dragged into
the converging clearance. High shear stresses drag the fluid into the wedge due to
the motion of the solid surfaces. In turn, high pressure must build up in the fluid
film before the viscous fluid escapes through the thin clearance. The pressure
wave in the fluid film results in a load-carrying capacity that supports the external
load on the bearing. In this way, the hydrodynamic film can completely separate
the sliding surfaces, and, thus, wear of the sliding surfaces is prevented. Under
steady conditions, the hydrodynamic load capacity, W , of a bearing is equal to the
external load, F, on the bearing, but it is acting in the opposite direction. The
U rh Uh
Re ¼ ¼ ð4-1Þ
m n
Here, h is the average magnitude of the variable film thickness, r is the fluid
density, m is the fluid viscosity, and n is the kinematic viscosity. The transition
from laminar to turbulent flow in hydrodynamic lubrication initiates at about
Re ¼ 1000, and the flow becomes completely turbulent at about Re ¼ 1600. The
Reynolds number at the transition can be lower if the bearing surfaces are rough
or in the presence of vibrations. In practice, there are always some vibrations in
rotating machinery.
In most practical bearings, the Reynolds number is sufficiently low,
resulting in laminar fluid film flow. An example problem is included in Chapter
5, where Re is calculated for various practical cases. That example shows that in
certain unique applications, such as where water is used as a lubricant (in certain
centrifugal pumps or in boats), the Reynolds number is quite high, resulting in
turbulent fluid film flow.
Classical hydrodynamic theory is based on the assumption of a linear
relation between the fluid stress and the strain-rate. Fluids that demonstrate such a
linear relationship are referred to as Newtonian fluids (see Chapter 2). For most
lubricants, including mineral oils, synthetic lubricants, air, and water, a linear
relationship between the stress and the strain-rate components is a very close
approximation. In addition, liquid lubricants are considered to be incompressible.
That is, they have a negligible change of volume under the usual pressures in
hydrodynamic lubrication.
Differential equations are used for theoretical modeling in various disci-
plines. These equations are usually simplified under certain conditions by
disregarding terms of a relatively lower order of magnitude. Order analysis of
the various terms of an equation, under specific conditions, is required for
determining the most significant terms, which capture the most important effects.
A term in an equation can be disregarded and omitted if it is lower by one or
several orders of magnitude in comparison to other terms in the same equation.
Dimensionless analysis is a useful tool for determining the relative orders of
magnitude of the terms in an equation. For example, in fluid dynamics, the
dt dx ¼ dp dy ð4-3Þ
Uh0
q¼ ð4-10Þ
2
The two boundary conditions of the velocity as well as the conservation of mass
condition form the following three equations, which can be solved for m, n and k:
0 ¼ m02 þ n0 þ k ) k ¼ 0
U ¼ mh2 þ nh
ðh ð4-11Þ
Uh0
¼ ðmy2 þ nyÞ dy
2 0
After solving for m, n, and k and substituting these values into Eq. (4-7), the
following equation for the velocity distribution is obtained:
1 h0 2 3h0 2
u ¼ 3U 2 3 y þ U y ð4-12Þ
h h h2 h
F IG. 4-3 Linear velocity distribution for a simple shear flow (no pressure gradient).
The two boundary conditions of the velocity and the constant flow-rate condition
form the three equations for solving for m, n, and k:
U ¼ m02 þ n0 þ k ) k¼U
2
0 ¼ mh þ nh þ U
ðh
Uh0
¼ ðmy2 þ ny þ U Þdy
2 0
After solving for m, n, and k and substituting these values into Eq. (4-7), the
following equation for the velocity distribution is obtained:
1 h 3h0 4
u ¼ 3U 2 03 y2 þ U yþU
h h h2 h
From the value of m, an identical expression to Eq. (4-13) for the pressure
gradient, dp=dx, is obtained for @h=@x < 0 (a converging slope in the x direction):
dp h h0 @h
¼ 6U m for < 0 ðnegative slopeÞ
dx h3 @x
This equation applies to a converging wedge where the coordinate x is in the
direction of a converging clearance. It means that the clearance reduces along x as
shown in Fig. 4-4.
In a converging clearance near x ¼ 0, the clearance slope is negative,
@h=@x < 0. This means that the pressure increases near x ¼ 0. At that point,
h > h0 , resulting in dp=dx > 0.
If we reverse the direction of the coordinate x, the expression for the
pressure gradient would have an opposite sign:
dp h h @h
¼ 6U m 0 3 for > 0 ðpositive slopeÞ
dx h @x
This equation applies to a plane-slider, as shown in Fig 4-5, where the coordinate
x is in the direction of increasing clearance. The unknown constant, h0 , will be
determined from the boundary conditions of the pressure wave.
4.6.2 Plane-Slider
In the case of an inclined slider, p0 is the atmospheric pressure. Pressure is
commonly measured with reference to atmospheric pressure (gauge pressure),
resulting in p0 ¼ 0 for an inclined slider.
The pressure wave, pðxÞ, can be solved for any bearing geometry, as long as
the film thickness, h ¼ hðxÞ, is known. The pressure wave can be solved by
analytical or numerical integration. The analytical integration of complex func-
hðxÞ ¼ ax ð4-15Þ
Here, x is measured from the point of intersection of the plane-slider and the
bearing surface. The minimum and maximum film thicknesses are h1 and h2,
respectively, as shown in Fig. 4-5.
In order to solve the pressure distribution in any converging fluid film, Eq.
(4-14) is integrated after substituting the value of h according to Eq. (4-15). After
integration, there are two unknowns: the constant h0 in Eq. (4-10) and the
constant of integration, po . The two unknown constants are solved for the two
boundary conditions of the pressure wave. At each end of the inclined plane, the
pressure is equal to the ambient (atmospheric) pressure, p ¼ 0. The boundary
conditions are:
at h ¼ h1 : p¼0
ð4-16Þ
at h ¼ h2 : p¼0
where b is the ratio of the maximum and minimum film thickness, h2 =h1 . A
similar derivation can be followed for nonflat sliders, such as in the case of a
slider having a parabolic surface in Problem 4-2, at the end of this chapter.
The velocity distribution can be substituted from Eq. (4-12), and after differ-
entiation of the velocity function according to Eq. (4-20), the shear stress at the
wall, y ¼ 0, is given by
3h0 2
tw ¼ mU ð4-21Þ
h2 h
Ff
f ¼ ð4-23Þ
W
dp @2 u
¼m 2 ð4-24Þ
dx @y
@u2 1 dp
¼ ð4-25Þ
@y2 m dx
The velocity profile is solved by a double integration. Integrating Eq. (4-25) twice
yields the expression for the velocity u:
1 dp 2
u¼ y þ ny þ k ð4-26Þ
2m dx
Here, n and k are integration constants obtained from the two boundary
conditions of the flow at the solid surfaces (no-slip condition).
The boundary conditions at the wall of the two plates are:
h0
at y ¼ : u¼0 ð4-27Þ
2
The flow is symmetrical, and the solution for n and k is
2
h
n ¼ 0; k ¼ m 0 ð4-28Þ
2
The parabolic velocity distribution is shown in Fig. 4-6. The pressure gradient is
obtained from the conservation of mass. For a parallel flow, there is no flow in the
z direction. For convenience, the y coordinate is measured from the center of the
clearance. The constant-volume flow rate, Q, is obtained by integrating the
velocity component, u, along the film thickness, as follows:
ð h=2
Q ¼ 2L u dy ð4-30Þ
0
dp 12m
¼ 3Q ð4-31Þ
dx bh0
This equation is useful for the hydrostatic bearing calculations in Chapter 10. The
negative sign means that a negative pressure slope in the x direction is required
for a flow in the same direction.
x2
hðxÞ ¼ hmin þ ð4-33aÞ
2R
dp h h
¼ 6mU 0 3
dx h
x20
h0 ðxÞ ¼ hmin þ ð4-33bÞ
2R
at x ¼ 1: p¼0
ð4-35Þ
at x ¼ 1: p¼0
For numerical integration, the infinity can be replaced by a relatively large finite
value, where pressure is very small and can be disregarded.
Remark: The result is an antisymmetrical pressure wave (on the two sides
of the minimum film thickness), and there will be no resultant load capacity. In
actual cases, the pressures are high, and there is a cavitation at the diverging side.
A solution that considers the cavitation with realistic boundary conditions is
presented in Chapter 6.
h
h ¼ ð4-37Þ
hmin
The equation for the variable clearance ratio as a function of the dimensionless
coordinate becomes
h ¼ 1 þ x 2 ð4-38Þ
h 0 ¼ 1 þ x 20 ð4-39Þ
In addition, if the dimensionless pressure is defined as
h2min 1
p ¼ pffiffiffiffiffiffiffiffiffiffiffiffiffi
ffi p ð4-40Þ
2Rhmin 6mU
Problems